Thanks -- I'll look for it.

I am being pretty careful -- especially in zeroing the cutter before starting each run. I can photo the sheet with the multiple different starting points. When it starts correctly, the motor is quiet, when it starts in a wrong place, there is a juddering noise when it accelerates in the X-direction. 2" thick by 8' long foam is pretty heavy, and it is like the stepper motor is getting forced and so it jumps.

 

I had hoped to try this evening, but I ran out of time. I will try as soon as I can -- I think the acceleration needs to be slowed, but I may be able to increase the feed rate a little. I'm going to redo the GCode for the 4mm (5/32") bit and I may set it for 3/16" to give a little cushion.

 

I tried the new settings last night, and it is much better, but the pauses in the X-axis still happen. I had the feed at 24"/minute and the 5/32" up-spiral bit handled it very well. It only got a bit out of sorts when it was cutting *full* depth because it had wandered; with a little cotton candy at the very top of the flutes.

It returned within about 3/4" on a 30" long part. Is the calibration process written out in a step by step process somewhere, or is it a trial and error process?

I cannot see the Sharpie attachment in the Phlatboyz store -- is it somewhere other than the accessories?

 

Can someone point me to a calibration FAQ, please?

 

Thanks Shaun -- I didn't see a calibration process on the DVD.

I changed out the tension wheels for bearings and it cuts the foam! Woohoo! They had a fair amount of friction, and it now cuts the big pieces (over 5' long) with less than 1/16" error!



I need to eek out a little more cutting depth it is a bit less than a 1/4" left and cutting it out leaves a rougher edge than I'd like -- the bit cut edges are smooooooooth. I also need a tweak or two on the layout -- and I'm cutting the next piece of foam "for real"!

 

I must have missed the memo. Which are the tension wheels?

 

Here's the bearings as tension idlers:



The good thing about the idler wheels is the timing belt has greater contact with the small drive pulley. I need all the torque I can get!

And here are the first 4 pieces sitting on top (in their approximate position) of the quarter scale model.



The original 24th scale and 12th scale egg crate models are on the right. These pieces are from my second test sheet, so the bigger pieces were all cut up from the earlier aborted starts, and you can see a few stray cuts on the right side passenger window piece. You can just make out the jigsaw puzzle joints at the apex of the sides and the roof.

Wish me luck tomorrow!

 

Neil, where did you get these tensioner?

 

I bought the bearings at my local Aubuchon hardware store. The OD was slight larger than the original wheels, and the ID was too big for the #10 motor screws, so I also got a bronze sleeve that closed the hole down enough, and the sleeve is longer than the two bearings, so I added the washers to hold everything in place.

The first "real" sheet cutting went *almost* well enough to use, but the sheet skewed a bit and two pieces ran off the edge, and there were some misaligned cuts by about the 5th piece (the big long one on the edge). The sheet was "walking" over to the left at the back. I stopped it before the last three pieces and realigned the sheet (though it then *almost ran off the back edge!).





I had to assemble it anyway to see the full scale of it. It looks big, but that may be because it is in the house. I'll take it outside tomorrow an see it in relation to my car.



I have to solve the cotton candy problem -- the relatively deep cuts bung up the bit when they don't line up, so I think I need to rig up an air nozzle to cool the bit and move out the shreds. I also need to get it cutting faster -- I had to slow it down to 25" per minute, so this took about 2 hours to finish. I was blowing it with my air brush, but I need a proper nozzle and a larger compressor.

I tested 42" / minute and 30" / minute and both started to cotton candy on the deep pass. So, I may need to find a bit with longer flutes/cutting surface, so the cuttings come up and out of the cut better, and/or I need the air nozzle. It sure would be nice to cut it faster -- this took about 2 hours plus.

Those are the battery bays at the bottom, by the way. And the thinnest parts are where the side windows will be - this is a partial height for them.

 

Hi,

I woke up this morning thinking about this and I'll try to put down my thoughts:

I need to cut *through* the foam -- it is actually 1 15/16" (50mm?) thick, so with my current up-spiral bit being 2 1/4" long, in theory I should be able to do this. I would have to add tabs, naturally, but this would help since it would lift the bit up once and a while.

I think the long sides of the 2' x 8' foam sheet needs to have guides. These could be ~3-4" diameter wheels with bearings -- these would need to be at least on the back of the machine but in the front as well would probably help. I will look for these ASAP.

Another much more involved method would be to have a light frame / tray to hold the sheet and to carry it straight and flat. My 7 support rollers are doing a reasonable job, but if we didn't need to depend on friction at the X-axis rollers to keep the sheet straight and had some sort of direct drive -- like a longer version of the Y-axis gantry belt, that would solve this issue, as well.

If the foam stays straight, and if some of the cutting dust can fall out the bottom, and the bit moves up and down some, and/or I limit the length of the cut to 2-3' at a time, then I can up the feed rate and still avoid cotton candy.

I will redo the layout of the F1 sheet to reduce the length of all edges to 2-3' long. Or, if it is possible to break up the lengths of the multi-path cuts: how can it cut 2' in one direction, then plunge in and go in reverse on that 2' length, and then continue. The object would be to automatically limit the time the bit is cutting deep in the groove, and to help clear out the dust, and to cool the bit by lifting it out of the groove.

If it is possible to separate the cutting path into segments, with tabs or by breaking them with a very short additional path, or to rewrite the way multi-path is programed, then I might not need to have an air nozzle blowing on the bit. I will see how short I can make the pieces and I'll add tabs and try to cut through the foam on the second pass -- and I see what sort of feed rates are then possible.

If anybody has thoughts on making an air nozzle (I'm thinking a small copper or aluminum tube that can be bent but then hold it's shape, with an air hose arching along side with the drive cable and vacuum hose, and I'll need a quieter and more powerful compressor.

Thanks in advance for any and all help / thoughts / hints! I'm certain this is doable, but the sheet must be keep from skewing and the X-axis motion must be steady (I'll be getting a more powerful stepper motor) and the bit must be kept cool enough to avoid cotton candy; either by limiting the length of individual cuts and/or with air flow.

 

I have problem with skewing too. I solved it with this contraption. On the right is the part that goes to the right side of the printer. I cut it out from MDF on the printer. On the left are just leftover pieces. I clamp it down with C-clamps. I used small angle aluminum but you can use 1" angle aluminum. The right part has to be fixed so that the part will be running true. The left part I set as needed by pushing it to the material but not to much pressure so that it will still glide smooth and easy. I use it for 7" material but you can adapt it to your needs.

Neil, you are an innovative kind of a guy or you wouldn't have started this project. Try cutting it in a different way. I don't know if it will work but you can do the experiment and report back to us. I have a piece of a hard steel. It used to be a shaft in an old printer. It's 3mm and doesn't bend easily. I think it can be used as hot-wire cutter. You can clamp it in the chuck of the flex-shaft and put a collar on top with one wire and have the other end of the rod stick bellow the drive rollers. Put another collar and a wire on that end. Make sure that the wire is long enough to travel from one end to the other without getting stuck. You'll have to experiment with the right setting of the current running through the rod and make sure your z-axis is disabled and that the g-code does not travel through parts but it you get it to work it will be the ideal case for your predicament with the thick foam.

Just my :02cents: and worth only what you paid for. Attached files

 

Someone here at work suggested using skateboard wheels as guides. I'll be working on it... I like the C-clamp ideal -- it would register every sheet in the same place, as well as limit the skewing.

I will try laying out smaller/shorter pieces and/or breaking the paths on the longer parts to force the bit to lift out and then continue. I'll see how it goes, but if you folks have any other thoughts, please post them!

 

If it was a shop vacuum, then I could; but I am using the central vacuum system. But maybe your suggestion means I go out and buy a nice (quiet?) shop vacuum?

 

Here's some more pictures of the F1 section outside next to my Scion xA for a sense of scale. (I've got aero-mods on the xA and I am averaging 41-55MPG per tank, year round.)



Any thoughts on sources of guide wheels with ball bearings, for the side rails I need to keep the foam from skewing?